Frank Hartwich et al. / Int. J. Food System Dynamics 3 (2010) 237-251
Table 1 shows that coffee growers are innovating mostly in agronomic aspects, such as improving
fertilization, shade and pest management. However, most farmers consider that they have not
implemented many changes in both production and improved quality aspects. The introduction of quality
standards and certification have been more prominent features in San Marcos and less so in El Pacon and
Las Crucitas. Over the past decade, San Marcos has become a region known for the cultivation of organic
coffee, so it is perhaps not surprising that the producers interviewed have acquired certifications or now
comply with different standards. Overall, farmers in San Marcos were slightly more innovative considering
the whole range of innovation aspects investigated.
4.3 Agents influencing innovation in coffee production: Financing institutions, buyers, development
agents, farmers organizations and input providers
Coffee growers were asked which agents have been helpful to inform them about measures for improved
coffee production. The type of agents and the innovation trajectories they lead have been discussed in
section 3. An overview on how many interactions with these agents farmers had in each of the
communities is presented. The results are depicted in Table 2.
Table 2 provides an indication of which innovation trajectories are most dominant in the community.
Whereas relationships to local buyers still dominate the situation in El Pacon, in Las Crucitas it is the
relationships with Development Agents and Input Providers that are most frequent. In San Marcos, it is
the relationships with the development agencies IHCAFE and ASONOG-FLO as well as with the cooperative
union AHPROCAFE and the local cooperative COCOSAM which are most influential in passing information
on innovations in coffee production. We may conclude that in El Pacon it is the buyer trajectory (a), in Las
Crucitas the input provider trajectory (d) and in San Marcos the mix of a development cooperation
trajectory (c) and farmers’ self initiative trajectory (b) that determine opportunities of absorption of
innovations for coffee growers. Drawing from the results on the average level of innovation in the
communities from the former section - San Marcos seems to have the most innovative growers - we may
ask ourselves if it is the latter mix which is most efficient in inducing innovations among farmers.
Table 2.
Number of interactions producers-agents
Type |
El Pacon |
Las Crucitas |
San |
Banks (BANCAFE, BANADESA, HSBC) |
1 |
5 |
0 |
Buyers (BELAGO, Beneficio Gaitan, Beneficio Sosa, Beneficio |
17 |
0 |
2 |
Development Agents (ACDI, CATIE, EDA, Fondo Cafetero, FUNDER, Heifer, ICADE, IHCAFE, SAG, FINTRAC, ASONOG-FLO) |
37 |
35 |
47 |
Farmers' Organizaitons (AHPROCAFE, COMIPIL, COCASAM)) |
3 |
6 |
42 |
Input Provider (Agrocomercial Gaitan, Agrocomercial Maribel, SEAGRO)_____________________________________________ |
3 |
17 |
0 |
Total |
61 |
63 |
91 |
4.4 Networks of information exchange
We collected data on the interaction between knowledge and technology providers and farmers (farmer-
to-farmer, and agent-to-farmer networks) which allowed us to construct knowledge and technology
exchange networks in the three communities. Growers were asked which growers and agents have been
helpful to inform them about measures for improved coffee production. Interviewees reported on their
relationship to a) coffee producers in their own community and b) agents in the coffee sector that passed
relevant information to them for improvement in coffee cultivation. In all the three communities the total
population of small coffee producers was interviewed, 25 in the community of El Pacon, 28 in Las Crucitas
and 25 in San Marcos. The resulting networks can be classified as full networks of the relationships
between coffee producers in each community combined with the cumulated egonetworks of the
producers with regard to their relationships with external agents.
The resulting dataset was analyzed with regard to general network properties and visualized using UCINET
software for social network analysis (Borgatti et al., 2002). We also plotted the networks using the spring-
embedding algorithm suggested in Net Draw (Borgatti et al., 2002). The size of the nodes reflects degree
centrality, viz the number of incoming ties of the actors; the colors indicate the type of actors (red for
245